|
Vintage Radio (domestic) Domestic vintage radio (wireless) receivers only. |
|
Thread Tools |
14th Jan 2022, 11:49 am | #41 | |||
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
Quote:
Looking at the datasheet for Brimar 12AH8, it has an equivalent noise resistance of RNv=100k ohms. The noise voltage is 18 microVolts at VHF. This seems massive compared with ECC85, RNv= 500ohms and ECF80 pentode section RNv=1500ohms. 100k may be the overall equivalent resistance of all sections. The noise figure RF is a non-linear function of RNv, it would give 12AHB a bad noise figure: https://www.sciencedirect.com/topics...nt-input-noise Quote:
This listing gives a rough snapshot of the types of FM valves used in different decades: https://mwsherman.com/fmonly/fm_only_list.html Note the prevalence of 12AT and 12DT8; I believe they are very similar except 12DT8 has an isolating earthed screen between triode sections. Note the 44-50MHz FM band in 1930s. Quote:
|
|||
15th Jan 2022, 1:10 am | #42 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: German one-valve superhet FM tuner
If you haven’t already seen it, BBC Research Report 1953-02 of 1953 January provides some information American and German FM/AM receivers of the time. It is available at: https://www.bbc.co.uk/rd/publications/rdreport_1953_02.
By then, the American receivers had settled down to their respective makers’ normal patterns. Zenith, RCA and GE all use the same front ends for both FM and AM, including the RF amplifier for AM. Zenith used a 6BJ6 RF amplifier with a 12AT7 mixer-oscillator. RCA used a 6CB6 RF amplifier with a 6X8 mixer oscillator. The mixer part was operated as a pentode on AM, but triode-strapped on FM. GE used a 6BJ6 RF amplifier, with a 12AT7 as mixer-oscillator on FM. On AM, the oscillator part of the 12AT7 was used, but the 1st 12AU6 pentode in the IF strip served as an AM mixer as well as 1st FM IF amplifier. (GE was also known to use pentodes in autodyne mode for AM frequency changing.) So both RCA and GE preferred the triode pentode combination for AM frequency changing. The German receivers all predated the availability of the ECC85 (later 1953, from Telefunken, I think.) Braun used an ECH42 frequency changer on both AM and FM, with an EF80 RF amplifier on FM only. Lowe Opta did similarly but with an ECH81 instead of the ECH42. Mende used an EF85 RF amplifier on FM only, then employed an ECH81 in a manner included in the original Philips information. That is, it was a regular frequency changer on AM, but on FM, the triode was used as an autodyne, with the heptode as 1st IF (presumably pentode-strapped). The next valve was also an ECH81, the heptode as IF amplifier on both bands (again, presumably pentode-strapped), the triode as an extra AF amplifier on AM (again something included in the original Philips notes.) Grundig used an ECH42 frequency changer without RF amplifier on all bands. My guess is that an appropriate RF amplifier ahead of a hexode/heptode mixer was acceptable for FM reception in reasonably strong signal areas. But without the RF amplifier, it might have been rather grim. The BBC (Spencer) simple FM receiver (Wireless World 1951 October & November) used a triode heptode frequency changer, initially an Osram X81, but later an X79 (which was said to be suitable for use up to 100 MHz). It had a Mazda 6F12 (= Z77, EF91) RF amplifier. Evidently the same basic circuit was adopted by Lowther for its first FM tuner (1953), but with different valves. The RF amplifier was a Brimar 6BW7 (perhaps directionally better than an EF80), the frequency changer was a Brimar 12AH8. The IF amplifier an EF80, and the limiter an EF42. A reasonable deduction is that the valves were chosen to extract the maximum performance from the basic circuit. Re the 12AH8, it was probably better described as being the best of a not-so-good bunch where FM was concerned. I did read somewhere (that I can no longer trace) that Brimar had traded off conversion gain for higher input resistance. That at least would explain the conversion gain of around 0.5 mA/V at a time when the “going rate” was around 0.75 mA/V (X79, ECH81, etc.) I suppose something to bear in mind that lowest possible noise was not always the goal with domestic receiver FM front ends, which would explain why some setmakers, like GE, preferred pentode over triode RF amplifiers until the end. A tuned input (itself with a noise debit) was easier with a pentode. The best answer was probably the cascode, but that was mostly reserved for better performance FM tuners in the USA, and quite rare in European FM practice. In the bipolar era, there was an RCA paper noting that a bandpass tuned input (four-gang circuit) gave noticeably improved spurious response rejection as opposed to a single-tuned input (three-gang circuit), but at the expense of 4 dB on the quieting curve (early part thereof), considered to be a worthwhile tradeoff. Cheers, |
15th Jan 2022, 12:32 pm | #43 |
Dekatron
Join Date: Mar 2005
Location: Gateshead, Tyne and Wear, UK.
Posts: 7,444
|
Re: German one-valve superhet FM tuner
From the September 1952 Wireless World.
Part 1 of the construction of a three valve FM feeder unit. Receiver is designed to cover the VHF broadcast band 87.5 to 100Mc/s. DFWB. |
15th Jan 2022, 12:50 pm | #44 | |||
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
Quote:
Quote:
Having double-tuned bandpass filiter with steep skirts of the response curve in the RF front end helps to filter out the spurious signals (intermodulation products). I have often seen the bandpass filters introduced in the mixer stage instead of the front end. It seems it is desirable not to have too much RF gain before it goes to the mixer, as the higher the gain, the more the Intermodulation products the mixer creates. Too much RF gain in the front end would introduce more Intermodulation junks or worst IP3 for closely spaced stations. Quote:
PS Attached photo is the Telefunken NT2 that I got last week. It is NOS in a factory box and working well. it has a bandpass filter in the mixer stage. Last edited by regenfreak; 15th Jan 2022 at 1:04 pm. |
|||
15th Jan 2022, 12:51 pm | #45 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
|
15th Jan 2022, 1:49 pm | #46 | |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
Quote:
|
|
16th Jan 2022, 11:22 pm | #47 | ||||
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: German one-valve superhet FM tuner
Quote:
Quote:
Returning to single-valve front ends, this is from a 1961 Zenith paper on stereo receiver circuitry: “Let us look at the RF tuner portion of this receiver. In Figure 1 we see a photograph of the complete tuner. This is a newly developed three gang permeability tuner featuring low noise reception with improved sensitivity. “Figure 2 contains the circuitry used in the tuner. A newly developed 6JK8 double triode tube is used utilizing frame grid construction for the RF section and a conventional triode for the autodyne converter section. Average noise figures for these tuners in production are 7Db measured in accordance with the IHFM standards. Image rejection ratios of 45Db are obtained.” That does suggest that Zenith had a hand in development of the 6JK8. Also, three gangs (including a tuned input) with a single-valve front end was unusual. But Zenith dropped back to two gangs, still using the 6JK8, for the production MJ1035. Quote:
Quote:
Cheers, |
||||
17th Jan 2022, 1:03 pm | #48 | |||
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
Many thanks!
Quote:
This is a very interesting design concept with 4-pole varactor tuned filter as shown in attachment 1! I have been experimenting with DIY 5-gang varactor FM tuner with 2-pole filter at the front end using a passive double-balance SBL-1 ring-mixer. The attachment 2 is my work in progress. It uses 6AK5 pentode as front end RF amp, ECC88 in cascode 2nd tuned RF, ECF80 as mixer and oscillator. I have encountered tracking and oscillator instability issues (sensitive to L/C ratio) as varicap is highly non-linear and lower Q than air gangs. Quote:
Quote:
|
|||
17th Jan 2022, 10:41 pm | #49 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
I have found these two unknown tuners in my storage box.
The two-valve tuner looks the same as this Russian Hi-Fi system with turntable: https://youtu.be/sOCQbp49DTQ The Youtube algorithm has recommended me to watch this video by pure coincidence. I was searching for its ID last week! In the video it shows the schematic. It uses 6N3 double triode in cascode which I have lots of experience with homebrewing FM tuners with it. The closet European counterpart would be ECC85. After the World War II, the US transferred lots of RCA valve technology to the USSR. So the russian radios were heavily influenced by the American designs. This tuner could be a clone of a Western design, I a not sure. The one-valve autodyne has Gorler and US patent number markings on it. Last edited by regenfreak; 17th Jan 2022 at 10:53 pm. |
1st Feb 2022, 11:10 pm | #50 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: German one-valve superhet FM tuner
[QUOTE=regenfreak;1441413]Attachment 3 is the heavy beast Sony CRF-230 that uses unusual double-tuned input. . It uses 5 -gang (2x RF, x x 1st IF, 2nd oscillation) with dual FM band 64-108MHz and dual conversion. Note the quirky 2-pole RF frond end using capcitive coupling network./QUOTE]
The Sony CRF-230 also appear to be unusual in having a bipolar cascode RF amplifier. At least I am not aware that this configuration was widely used. The bipolar cascode was sometimes found in TV IF amplifiers, and as a mixer in some American VHF TV tuners. (RCA used it following a Nuvistor RF amplifier, later using a dual-gate mosfet RF amplifier.) On the other hand, the jfet cascode was used as an FM RF amplifier in the interval between the bipolar era and the advent of dual protected gate mosfets. (And I think Heathkit stayed with that arrangement well into the 1970s.) The dual-gate mosfet RF amplifier was inherently a cascode. When used as an FM mixer, the typical arrangement was what might be called a “half-cascode”, in that it was a cascode to the signal input on g1, but not to the oscillator input on g2. (VHF TV applications used a full cascode, with oscillator and signal both on g1 to avoid the low-Band I channel regeneration issue, and in the early days, TI suggested the same for FM.) Valve cascode RF amplifiers were also used for FM, soon after their adoption for VHF TV, in both series and shunt forms, and with both “regular” and frame-grid valves. Returning to solid-sate, the mixed cascode of singe-gate mosfet and jfet was also known, see attachment. The single-gate mosfet seemed to have not so many applications, although well-know is Leak’s use of such, solo, as the RF amplifier in its Stereofetic FM tuner. I wonder if that choice was simply a case of where it chose to get off the RCA bus. RCA had moved to mosfets in stages in its published FM front-end circuitry: from all bipolar to single-gate mosfet RF with bipolar mixer; to single-gate mosfets in both positions; to single-gate mosfet RF and dual-gate mosfet mixer; to dual-gate mosfets in both positions. Cheers, |
2nd Feb 2022, 12:46 am | #51 | |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
Quote:
In the 2nd attachment, it is the response curves of three bandpass for a german car radio of unknown model that I found on internet. PS I am not familiar with TV tuners but this page gives very good historical overview of their evolution in designs. It talks about noise figure as well https://www.maximus-randd.com/tv-tuner-history-pt2.html Last edited by regenfreak; 2nd Feb 2022 at 1:01 am. |
|
8th Feb 2022, 2:05 am | #52 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: German one-valve superhet FM tuner
Something of a surprise find was an RCA circuit for a cascode Nuvistor front end, attached. This was in the RCA 1964 Receiving Tube manual, RC-23.
When the Nuvistor was introduced, RCA used a single 6CW4 as the RF amplifier in a VHF TV front end, claiming that it gave superior overall performance to other forms, including the frame-grid cascode, shielded grid triode (including the frame-grid variant) and low-screen-current tetrode. (See RCA Engineer 1960 August, September p.28ff.) Thus a cascode Nuvistor seems almost “over-the-top”. Whether it was used in a production tuner I do not know. But I recall that the single-Nuvistor RF amplifier was used by one or other of the US tuner makers, perhaps Scott. At the other end of the spectrum was another RCA FM tuner circuit, which remained in its Receiving Tube Manuals at least until the 1961 edition, RC-21. With its 6BE6 mixer, albeit externally excited, it looks like a relic from the 1940s. The 6BJ6 dated from late 1946. And if we assume – reasonably I think – that RCA would have used the 6BA7 as mixer once it became available during 1948, then that puts it at around 1947. And from c.1952, RCA’s “norm” seemed to be a 6CB6 RF amplifier with a 6X8 frequency changer. The 6BJ6 was an interesting valve. It was derived from the 6BA6, but with some “tweaking” the heater requirement was halved, to 150 mA, and the VHF (100 MHz) performance was significantly improved. In part it was aimed at the then emerging VHF mobile market, where low power consumption was an attribute. But its heater characteristics also meant that the same valve could be used in both 6.3-volt and 150 mA series-string circuits. And in the latter case, its low voltage requirement (as compared with the 12.6 V or more required for most 150 mA valves) was an asset in American AC-DC AM-FM receivers, where typically seven valves had to be fitted into the available 117 volts. The GE FM tuner circuit is from its Essential Characteristics tube data book, the latter undated but I’d say that the circuit was from the early-to-mid 1950s. (The 6BZ7 cascode valve dates from 1952.) It was probably reasonably representative of cascode FM RF amplifier practice of the period. The GE AC-DC AM-FM receiver circuit was also reasonably representative of lower-level American practice before the single-valve front end arrived, with a 6BJ6 RF amplifier and a 12AT7 mixer-oscillator. Cheers, |
8th Feb 2022, 9:08 pm | #53 | ||
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
Quote:
I do not know if nuvistor cascode is way more superior than framegrid ECC88. But ECC88 has higher noise figure with Rn = 300 ohms. In RCA Engineer 1960 August, September, It compared Rn of 6CW4 (=181 ohms) with 6ER5 and 6FH5 which i never heard of. Surely ECC85 cascode can be found in some tuners and they are not as good as nuvistor in cascode. But ECC88 and 6BS8 were designed to be used in cascode. I believe 6BZ7 is not the same 6SB8. I tried replacing ECC85 with 6SB6 in my homebrew as ocillator, its output is more uniform across the frequency range. Quote:
Last edited by regenfreak; 8th Feb 2022 at 9:38 pm. |
||
25th Mar 2022, 3:37 am | #54 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: German one-valve superhet FM tuner
A 1953 September Radio-Electronics article on cascode fronts ends (for TV and FM) included (incorrectly, I think) one FM front end that in fact used a pair of grounded grid stages in cascade. Probably because it was included in the cascode groups, there was no commentary as to how it compared with the regular series cascode circuit.
Nonetheless, the double grounded grid RF amplifier did turn up again, in a 1957 RCA circuit for a VHF TV tuner with constant input impedance. It does not seem to have had much influence on actual VHF TV tuner practice, though. Soon after, there was a move to the use of specialized single triodes rather than cascode double triodes as RF amplifiers, to some extent on the basis that the cost decrement so obtained was bigger than the performance decrement – the old “nearly as good” premise. Cheers, |
25th Mar 2022, 8:17 pm | #55 | |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: German one-valve superhet FM tuner
Quote:
In the other thread (https://www.vintage-radio.net/forum/...88#post1458288 ) , i was going to use a pair of 6cw4 in cascode because of the much lower noise figure (see attachments). But 6cw4 are too pricy. So I ordered four Soviet military spec 6s52n-v from Ukraine in January, They are supposed to be identical to 6Cw4. Then I changed my mind again, opting for more simple construction and replaced two nuvistors by one E88CC in cascode. The Marantz 10B uses two-stage cascade EC88 single triodes (not the same as E88CC or ECC88) amps with grounded grids (see attachment). Although it has 8-gang air capacitors, effectively the 10B has 3-tuned pi-networks cascade Rf amp to compensate the insertion loss of the passive ring mixer. (All of the Marantz design team members did not have a university degree in electrical or electronics engineering.) Last edited by regenfreak; 25th Mar 2022 at 8:30 pm. |
|